Increasingly, studies suggest benefits of natural environments or greenness on children's health. However, little is known about cumulative exposure or windows of susceptibility to greenness exposure. Using inverse probability weighting of marginal structural models (IPW/MSM), we estimated effects of greenness exposure from birth through adolescence on executive function and behavior. We analyzed data of 908 children from Project Viva enrolled at birth in 1999–2002 and followed up until early adolescence. In mid-childhood (median 7.7 years) and early adolescence (13.1 years), executive function and behavior were assessed using the Behavior Rating Inventory of Executive Function and the Strengths and Difficulties Questionnaire (SDQ). Greenness was measured at birth, early childhood, mid-childhood, and early adolescence, using the Normalized Difference Vegetation Index. We used inverse probability weighting of marginal structural models to estimate effects of interventions that ensure maximum greenness exposure versus minimum through all intervals; and that ensure maximum greenness only in early childhood (vs. minimum through all intervals). Results of the effects of “maximum (vs. minimum) greenness at all timepoints” did not suggest associations with mid-childhood outcomes. Estimates of “maximum greenness only in early childhood (vs. minimum)” suggested a beneficial association with mid-childhood SDQ (−3.21, 99 %CI: −6.71,0.29 mother-rated; −4.02, 99 %CI: −7.87,-0.17 teacher-rated). No associations were observed with early adolescent outcomes. Our results for “persistent” maximum greenness exposure on behavior, were not conclusive with confidence intervals containing the null. The results for maximum greenness “only in early childhood” may shed light on sensitive periods of greenness exposure for behavior regulation.

越来越多的研究表明自然环境或绿色环境对儿童健康有益。然而，关于累积暴露或对绿色暴露的敏感性窗口知之甚少。使用边际结构模型 (IPW/MSM) 的逆概率加权，我们估计了从出生到青春期接触绿色环境对执行功能和行为的影响。我们分析了来自 Project Viva 的 908 名儿童的数据，这些儿童在 1999 年至 2002 年间出生时就注册了，并一直随访到青春期早期。在儿童中期（中位数 7.7 岁）和青春期早期（13.1 岁），使用执行功能行为评级量表和优势与困难问卷 (SDQ) 评估执行功能和行为。绿色度是在出生、儿童早期、儿童中期和青春期早期测量的，使用归一化差异植被指数。我们使用边际结构模型的逆概率加权来估计干预措施的效果，以确保在所有时间间隔内确保最大绿色暴露与最小绿色暴露；并确保仅在幼儿期达到最大程度的绿色（相对于所有时间间隔内的最小程度）。“所有时间点的最大（相对于最小）绿色”的影响结果并未表明与儿童中期的结果有关。对“仅在幼儿期的最大绿色度（相对于最小值）”的估计表明与儿童中期 SDQ 有益相关（−3.21, 99 %CI: −6.71,0.29 mother-rated; −4.02, 99 %CI: −7.87, -0.17 教师评分）。没有观察到与早期青春期结果的关联。我们对行为的“持续”最大绿色暴露的结果，对包含空值的置信区间不是决定性的。“仅在儿童早期”的最大绿色度的结果可能会阐明行为调节的绿色暴露敏感期。